[osis-core] OSIS pointers and references

[Patrick Durusau]

Harry,

Sorry for the delayed response! Got pulled off into Hebrew display in 
Flash in a Mac development environment! Nasty stuff!

Harry Plantinga wrote:

>>I do not think of the work element as only referring to an OSIS
>>document. <work> is (in your words) "a general bibliography item..."
>>
>>What I envisioned for <work> was as follows:
>>
>><work> with an attribute value of workID could record as any general
>>bibliography item, any work, from electronic to print to ancient.
>>
>>The workID value creates a mapping between the item as recording as a
>>general bibliographic reference and the workID.
>>
>
>Agreed.  Then are we overloading a <work> element when we use it to map
>"lxx" to "any English translation of the LXX"? Is saying that "osisRefs
>prefixed by lxx: come from the osisIDScheme 'bible.lxx.en'" really making a
>bibliography entry?
>
>Maybe the aswer to that would be the same as the answer to this question:
>Does using a <work> to declare "lxx" to mean "any English translation of the
>LXX", necessarily imply that "any English translation of the LXX" should go
>into the bibliography?
>
Yes.

In other words, the bibliography entry (<work>) that declares the workID 
bible.lxx, may not refer to an actual edition, but could refer to the 
commonly known LXX reference scheme. It could also refer to a particular 
edition, which would be useful if you wanted to be very specific about a 
particular reference system that was being identified by the workID.

Hmmm, I am trying to say that the bibliographic entry (<work>) may 
identify a reference system without pointing to any particular example 
of that system. Such as bible.nrsva would point to the NRSVA system, 
without necessarily having to point to the NIV or other particular 
edition. If I used bible.nrsva, what I anticipate to be the behavior 
would be that any work using the bible.nrsva reference system would be a 
legitimate target of a reference/pointer from this document.

Assuming that I declare in a particular OSIS document instance that it 
is following the bible.nrsva reference system, then all 
references/pointers are by default using that system.

Admittedly a little different from normal bibliography entries that 
always point to a particular work. This does allow us to specify the 
origin of a system (NA27 with full citation).

I think we probably should illustrate in prose the appropriate level of 
detail that should be expected and give some common samples that people 
can paste into their documents.

(Noting that the reference declaration(refsDecl in TEI), the thing Harry 
asked about that specifies how to validate a particular reference 
system, is the most important item for a separate document after OSIS 
1.1 appears.)

Not sure if I would say we are overloading the <work> element, since 
several may appear, but only the one that is used by <osisText> has the 
behavior of specifying the default reference system for the document. 
(Multiple <work> elements, only one attribute on osisText to identify 
the reference system for the document.)
We should strongly urge that any wordID must have an entry in a <work> 
element. (And as I said, we can offer a common set for regular use.)

(The default behavior does not limit the number of reference systems 
that can be used within a document instance. It merely provides a 
default system that can be overriden at any point.)

Closer?

Patrick





>
>-Harry
>

-- 
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
pdurusau@emory.edu